| xxiao | i can do python setup.py fine on x86, but it always gave me 'illegal instructions' on ppc, am I missing something on python build env? | 02:41 |
|---|---|---|
| Chipzz | logout | 02:45 |
| Chipzz | *facepalm* my bad, ignore pls :P | 02:45 |
| infinity | xxiao: If you can sort out what the illegal insn is... Python could be miscompiled (perhaps has AltiVec enabled when it shouldn't) | 02:51 |
| infinity | xxiao: All our buildds support AltiVec, so we don't catch that in production when it slips in. :/ | 02:52 |
| xxiao | infinity: on this particular chip i actually had altivec | 02:53 |
| xxiao | yes i'm trying to chase it down | 02:53 |
| infinity | xxiao: Oh, kay. I knew some of the fsl chips BenC was working with didn't have Altivec, so that was my shot in the dark. | 02:54 |
| infinity | xxiao: Seems unlikely that python is just plain fundamentally broken on PPC, or it would break a huge number of builds in the distro. | 02:54 |
| xxiao | so far i had no idea on how to debug python setup.py breakage, googling | 02:55 |
| xxiao | i wrote a simple setup.py and it worked | 02:55 |
| xxiao | so probably some 'import' caused this trouble | 02:55 |
| xxiao | i bootstrapped the rootfs from 12.10 ubunut-core for ppc | 02:55 |
| xxiao | the kernel is 64bit | 02:56 |
| xxiao | the python setup.py died immediately, i.e. even 'python setup.py help' will not run | 02:56 |
| xxiao | from sphinx.setup_command import BuildDoc Illegal instruction | 03:07 |
| xxiao | seems like the illegal instruction has something to do with BuildDoc and sphinx | 03:07 |
| xxiao | maybe I'm missing some sphinx pkgs? | 03:08 |
| xxiao | everytime i saw documentation related errors from debian/ubuntu it scares me | 03:11 |
| xxiao | now I can reproduct this easily: put one liner to a python script: | 03:14 |
| xxiao | from sphinx.setup_command import BuildDoc | 03:14 |
| xxiao | then run it, it will illegal instruction | 03:14 |
| xxiao | s/reproduct/reproduce/ | 03:15 |
| xxiao | ahh...i could not run sphinx at all, "sphinx-quickstart' will crash on ppc right away | 03:16 |
| infinity | xxiao: Which version of python-sphinx is that? The one from quantal or raring? | 03:18 |
| xxiao | quantal | 03:18 |
| infinity | Although, hrm. Can't be sphinx's fault, it's arch:all. | 03:18 |
| * infinity spins up a PPC machine to test this. | 03:19 | |
| xxiao | thanks | 03:20 |
| infinity | xxiao: So, you say that 'sphinx-quickstart' just crashes instantly for you? | 03:23 |
| infinity | xxiao: Works fine on my POWER5 here, on both raring and quantal. | 03:23 |
| infinity | xxiao: http://paste.ubuntu.com/5560511/ | 03:24 |
| xxiao | correct, it crashes on my t4240 box | 03:24 |
| infinity | Or do I need to fill in some values, does it crash later? | 03:25 |
| infinity | Hrm, no, seemed to go all the way through fine. | 03:26 |
| xxiao | http://paste.ubuntu.com/5560515/ | 03:26 |
| infinity | Well, your kernel seems to be living in the distant past, but that's probably not the issue. | 03:27 |
| xxiao | we're upgrading 3.0 to 3.8 these days, will math-emulation matter? | 03:27 |
| infinity | Possibly. BenC probably knows the answer to that better than I do. | 03:28 |
| infinity | I'm really unfamiliar with the fsl cores. | 03:28 |
| xxiao | it seems t4240's FPU has one special instruction that diffs from power5 | 03:28 |
| infinity | My house is full of old Motorola cores, and new IBM ones. | 03:28 |
| xxiao | i c, will talk with BenC then | 03:28 |
| xxiao | meanwhile I will try math-emu | 03:28 |
| xxiao | thanks for the help | 03:29 |
| infinity | Let me try this on an old PPC750 | 03:29 |
| xxiao | I assume you have math-emu on? | 03:30 |
| infinity | I have... Whatever our distro kernels have. | 03:31 |
| xxiao | i'm bulding kernel with math-emu | 03:34 |
| xxiao | had to do this with yocto, which means 30 minutes to get a clean build, sigh | 03:34 |
| BenC | xxiao: Shouldn't matter | 03:34 |
| BenC | xxiao: You're going to be in Austin next weekend, right? | 03:34 |
| xxiao | BenC: yes | 03:34 |
| BenC | Flying out there tomorrow | 03:34 |
| xxiao | that's early | 03:34 |
| xxiao | but...welcome to Austin | 03:34 |
| BenC | Going to Dallas on Monday (Servergy office) and coming back to Austin later on in the week | 03:35 |
| xxiao | whenever i saw doc-related crash, i thought about fpu, which is troublesome for fsl chips | 03:35 |
| infinity | BenC: Want to try to sort out this illegal insn issue xxiao is seeing? I can't reproduce here, and I don't have the right hardware to dig deeper. | 03:35 |
| xxiao | we will have weeklies with servergy | 03:35 |
| BenC | just cheaper to round trip to Austin (and drive to/from Dallas) rather than make a 3-point trip | 03:35 |
| BenC | infinity, xxiao: Can you give me a repro case? | 03:36 |
| infinity | BenC: Install python-sphinx, and run sphinx-quickstart | 03:36 |
| infinity | BenC: Possibly quantal-specific. | 03:36 |
| infinity | xxiao: Did you try on raring? | 03:36 |
| xxiao | infinity: not yet | 03:36 |
| xxiao | i will try it on raring tomorrow | 03:37 |
| infinity | BenC: Can't reproduce on any of the IBM cores I throw it at. | 03:38 |
| BenC | infinity: should crash on start? | 03:38 |
| infinity | (Well, I'm still debootstrapping to test on my 750, but I imagine it'll be fine) | 03:38 |
| xxiao | BenC: or one liner python : from sphinx.setup_command import BuildDoc will trigger it | 03:38 |
| BenC | raring doesn't seem to show it | 03:38 |
| infinity | BenC: Crashes at startup, yeah. | 03:38 |
| BenC | On e500mc that is | 03:38 |
| infinity | BenC: http://paste.ubuntu.com/5560515/ | 03:38 |
| BenC | Checking quantal | 03:39 |
| infinity | BenC: Then again, I'm willing to blame xxiao's sketchy kernel there, too. ;) | 03:39 |
| BenC | xxiao: what does gdb show as the illegal instruction? | 03:39 |
| xxiao | hold on | 03:39 |
| BenC | Doesn't crash on e500mc on quantal | 03:40 |
| infinity | Works fine on a 15-year-old PPC750 too. | 03:43 |
| xxiao | crappy t4240 then | 03:44 |
| xxiao | sphinx* is all pythin scripts, don't know how to gdb the instruction | 03:44 |
| xxiao | it happens when "from sphinx.something import somefunc" | 03:45 |
| BenC | xxiao: gdb python | 03:45 |
| BenC | xxiao: set args -f sphinx-quickstart | 03:45 |
| xxiao | http://paste.ubuntu.com/5560553/ | 03:47 |
| infinity | *blink( | 03:47 |
| infinity | Under gdb, I get a SIGILL. | 03:47 |
| infinity | Double-U Tee Eff. | 03:47 |
| xxiao | __sqrt_finite () | 03:47 |
| xxiao | I KNEW it! | 03:47 |
| xxiao | last time a floating related PHP bug on ppc haunted me for two days | 03:47 |
| xxiao | i ended up using integer to work around that, instead of compiling all those php stuff from source | 03:49 |
| infinity | xxiao: Is it dying in OPENSSL_cpuid_setup() for you, or elsewhere? | 03:51 |
| xxiao | the pastebin is all I saw | 03:52 |
| xxiao | why is openssl involved here, curious | 03:52 |
| infinity | Oh, mine's a false alarm. | 03:52 |
| infinity | OPENSSL_cpuid_setup just can't trap SIGILL when run under gdb. | 03:53 |
| infinity | xxiao: Nevermind me, I'm just being a doofus. | 03:53 |
| infinity | xxiao: Your crash is elsewhere, and more interesting. :P | 03:53 |
| infinity | xxiao: disassemble 0x0fe9aa94 | 03:56 |
| xxiao | http://paste.ubuntu.com/5560579/ | 04:06 |
| Bluefoxicy | ffs I need to replace the wall in this room with a big white board. | 04:07 |
| xxiao | BenC: math-emu seems helped | 04:16 |
| xxiao | infinity: with math-emu now it's ok | 04:17 |
| xxiao | is there nodejs for ppc? | 04:38 |
| BenC | xxiao: nope…v8 needs a lot of porting for ppc | 04:38 |
| infinity | Yeah. :/ | 04:38 |
| infinity | There's a lot of v8-using stuff out there these days too. | 04:39 |
| xxiao | BenC: i'm 'recommending" freescale to become openstack member asap | 04:39 |
| BenC | xxiao: We were going to bring up v8 with Freescale this week :) | 04:39 |
| infinity | BenC: I stumbled on a v8/ppc "port" on github a while back, but I suspect it stalled before it ever really got started. | 04:39 |
| xxiao | the openstack guys mandated nodejs for its dashboard, that kills ppc/mips | 04:39 |
| BenC | infinity: It's completely empty from > 1 year ago | 04:39 |
| infinity | BenC: I'd love if someone just wrote a *&%! generic C implementation of v8, to be honest. | 04:39 |
| infinity | BenC: Sure, it might be slow, but at least it would work everywhere. | 04:40 |
| BenC | infinity: amen... | 04:40 |
| infinity | xxiao: There's a MIPS v8 port that actually works, AFAIK, it's just not upstream. | 04:40 |
| infinity | xxiao: But yeah, nothing for PPC. | 04:40 |
| BenC | xxiao: That's fantastic…openstack is right where we want to position CTS-1000 hardware | 04:40 |
| infinity | BenC: If you can talk anyone into putting resources into v8/ppc, that'd be awesome. If you can talk anyone into generic C, I won't complain. :P | 04:41 |
| BenC | infinity: I have a feeling I'll be buying lots of jaeger bombs this week…let's see if it gets things moving | 04:42 |
| * infinity can't help but think this is the sort of thing IBM would have thrown money/people at 10 years ago, but trying to convince them of it today would be hard. | 04:42 | |
| infinity | But perhaps worth asking benh, on the off chance they've been doing something there and just not been vocal. | 04:42 |
| xxiao | ok lxc/12.10 is up | 04:44 |
| xxiao | time to tie it with openstack via cli, as no dashboard can run without nodejs | 04:44 |
| Bluefoxicy | infinity: no way, Websphere is infinitely superior to nodejs, and you should be using it instead. Ask IBM. | 04:45 |
| infinity | Bluefoxicy: v8's embedded in a lot of products other than nodejs. Lacking the port sucks period. | 04:46 |
| Bluefoxicy | infinity: Also, MySQL sucks terribly, but if you need something better there's this database called Oracle that the folks who currently control MySQL would be happy to help you deploy... | 04:46 |
| infinity | (And lacking a generic implementation sucks for new arches) | 04:46 |
| erickLee | hi | 04:47 |
| BenC | xxiao: openstack-dashboard works for me | 04:48 |
| erickLee | can anyone tell me what is the safe package to attempt modifications? | 04:48 |
| xxiao | http://paste.ubuntu.com/5560653/ | 04:48 |
| BenC | Although, I think I used it on x86 | 04:48 |
| xxiao | what the hack, it takes 2 minutes for lxc to get a login prompt | 04:48 |
| Bluefoxicy | infinity: I suffered major stage 12 burnout recently and have come out incredibly, terribly critical and horrible somehow. I make no excuses or apologies for this, but keep your eyes peeled for cynicism. | 04:48 |
| xxiao | BenC: which openstack version? they mandated nodejs for folsom | 04:49 |
| xxiao | if you're doing Essex it might be fine | 04:49 |
| BenC | xxiao: I don't know the versions, but that's from raring | 04:49 |
| xxiao | BenC: hmm need try that then | 04:49 |
| xxiao | BenC: i'm building the whole thing from source, maybe I should do deb install instead | 04:50 |
| BenC | xxiao: Might save you some time :) | 04:50 |
| xxiao | build it from source is easy to hack in lxc support, which is not officially supported | 04:50 |
| infinity | nova-compute-lxc is packaged. | 04:51 |
| infinity | You shouldn't need to build from source for that. | 04:51 |
| xxiao | yeah will try the package install tomorrow | 04:51 |
| infinity | It's how we currently do openstack on ARM, due to the lack of paravirt solutions there. | 04:52 |
| xxiao | i spent the whole day to get it fully built from devstack, just missing nodejs | 04:52 |
| vibhav | You can get away with misaligned access on x86, right? | 04:52 |
| xxiao | not sure how ubuntu bypassed that | 04:52 |
| infinity | vibhav: Yes, but don't. | 04:52 |
| * vibhav grumbles | 04:53 | |
| infinity | vibhav: For two reasons. (a) it's bad practice that creates unportable code and (b) it's slower, cause you waste cycles on fixups. | 04:53 |
| vibhav | infinity: Indeed. Thoggen had failed to build on arm for this reason | 04:54 |
| vibhav | And I was finding ways to fix it | 04:54 |
| xxiao | BenC: who can I ask further on openstack/raring/ppc? | 04:56 |
| xxiao | esp the openstack version and nodejs dependency | 04:56 |
| lifeless | xxiao: whats wrong with nodejs ? | 04:58 |
| BenC | xxiao: If there was a nodejs dep, I would not be able to install it :) | 04:58 |
| BenC | lifeless: it doesn't run on ppc | 04:58 |
| BenC | xxiao: *wouldn't | 04:58 |
| BenC | err, had it right I guess | 04:58 |
| xxiao | ok will try install from pkgs tomorrow | 05:03 |
| xxiao | http://packages.ubuntu.com/quantal/python-django-horizon shows no nodejs | 05:03 |
| xxiao | http://changelogs.ubuntu.com/changelogs/pool/main/h/horizon/horizon_2012.2-0ubuntu2/changelog | 05:04 |
| xxiao | Drop node-less dependency (LP: #1024326) | 05:04 |
| ubottu | Launchpad bug 1024326 in horizon (Ubuntu Quantal) "node.js is required for access to the dashboard" [High,Fix released] https://launchpad.net/bugs/1024326 | 05:04 |
| xxiao | that is it I think | 05:04 |
| lifeless | BenC: ah; shallow or deep issues do you think ? | 05:05 |
| BenC | lifeless: deep…v8 needs to be ported | 05:11 |
| BenC | xxiao: right, but it jut means you need nodejs to access it, not to run it, right? | 05:11 |
| BenC | I'm wrong, looks like it needs it on the host system | 05:12 |
| BenC | xxiao: But it does look like it's fixed in quantal and raring | 05:14 |
| lifeless | so the nodejs dep | 05:14 |
| xxiao | BenC: yes | 05:14 |
| lifeless | is currently just as a compiler for lesscss | 05:14 |
| lifeless | so you can run that once per revision on any architecture and then its done. | 05:15 |
| xxiao | BenC: basically it tells openstack don't do dynamic compilation, use a pre-built css instead | 05:15 |
| xxiao | i'm pulling that into devstack now, will see how it goes | 05:15 |
| lifeless | however its likely upstream will want to do much more dynamic things later, so I wouldn't bet on it staying as an indefinite workaround | 05:15 |
| xxiao | lifeless: that's why I want to ask freescale to get on openstack paid membership to vote against this tech decision | 05:16 |
| xxiao | or port v8/nodejs to ppc | 05:16 |
| lifeless | paid membership gets a board position; nodejs is a tech decision that the board doesn't vote on | 05:16 |
| xxiao | requires nodejs just for the sake of css seems odd to me | 05:17 |
| lifeless | the way to get technical leverage is to contribute | 05:17 |
| lifeless | technically | 05:17 |
| lifeless | you could write a python LESS compiler | 05:17 |
| lifeless | then get the nodejs dep dropped, as long as it has not expanded in scope | 05:18 |
| === doko_ is now known as doko | ||
| === zequence_ is now known as zequence | ||
| === highvolt1ge is now known as highvoltage | ||
| === Zilvador_ is now known as Zilvador | ||
| jbicha | hi, could we get a rebuild of cmake, gvfs, and packagekit for the libarchive transition? | 15:10 |
| Laney | jbicha: what about all the others? | 19:40 |
| infinity | Laney: I'm working on the whole stack right now. | 19:43 |
| infinity | Laney: Though cmake might be stuck on an FTBFS in the testsuite. *sigh* | 19:44 |
| jbicha | infinity: thanks | 19:45 |
| * infinity scratches his head at cmake's newfound hatred for its testsuite. | 19:46 | |
| xxiao | I was told turning on math-emu is the wrong thing to do, not sure if that's a correct suggestion | 20:16 |
| xxiao | infinity: for ppc of course, if kernel has math-emu off, will the sphinx-quickstart abort? | 20:17 |
| xxiao | I would assume ubuntu rootfs is soft-float by default? | 20:17 |
| xxiao | BenC: pinh | 20:25 |
| xxiao | ping | 20:25 |
| infinity | xxiao: Why would you assume Ubuntu is soft-float? | 20:25 |
| xxiao | infinity: just guessing, is it? | 20:26 |
| infinity | Isn't soft-float on PPC a reasonable new (and incompatible) ABI? | 20:26 |
| xxiao | infinity: in my case, ppc rootfs in embedded space is mostly soft-float | 20:26 |
| xxiao | esp those e500 chips | 20:27 |
| xxiao | if everything is hard float, then why we need math-emu? | 20:28 |
| infinity | I'm not familiar with the math-emu you keep talking about. | 20:29 |
| infinity | Is this specific to fsl kernels? | 20:29 |
| infinity | If so, it's probably trapping instructions that don't exist on your cores. | 20:29 |
| xxiao | math-emu is a kernel option that traps floating point instruction if there is no fpu | 20:40 |
| === Ursinha_ is now known as Ursinha | ||
| === Ursinha is now known as Ursinha-afk | ||
| === Ursinha-afk is now known as Ursinha | ||
| === Ursinha is now known as Ursinha-afk | ||
| === Ursinha-afk is now known as Ursinha | ||
| wooo | What is vnode? Is is somewhat similar to inode? | 23:39 |
| gkcn | wooo, http://www.linuxquestions.org/questions/linux-kernel-70/difference-between-inode-and-vnode-657954/ | 23:45 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!